PROJECT 2 - Identifying Sources Linked to Differential Assertive Community Treatment (ACT) Outcomes for Latinos In this research we propose to conduct statistical analyses of data from a large quasi-experimental project (ACCESS) and a 2-stage modified Delphi technique to generate recommendations for the modification of assertive community treatment (ACT) when applied to Latino and other severely mentally ill homeless (SMIH) minority populations. The ACCESS study compared the outcome effects of "augmented" ACT (i.e., ACT + quality improvement intervention) and standard ACT, and found no evidence of differences. Utilizing a disparities framework, advanced statistical methods, and a 2-stage Delphi method, we propose to conduct service and outcomes disparities research aimed at generating information with clinical or programmatic utility for Latinos. The proposed analyses will generate new evidence on service use and outcomes for Latino SMIH persons treated with ACT, and will broaden our understanding of the effects of quality improvement interventions for Latinos. Findings will be integrated with expert opinion to generate recommendations regarding ACT for Latino SMIH persons. The study has four aims: 1. To test for disparities in service use for Latino and African American as compared to white SMIH persons served by standard and augmented ACT 2. To test for outcome disparities for Latino and African American as compared to white SMIH persons served by standard and augmented ACT 3. To estimate the impact of observed service disparities on observed outcome disparities for Latino and African American as compared to white SMIH persons 4. To integrate results from analyses of service and outcome disparities for Latinos, a literature review, and expert opinion, and generate recommendations to improve ACT for Latino SMIH persons.